Fix with OpsAI

Fix with OpsAI investigates production issues using your observability data and prepares proposed solutions. It connects the observed error with available telemetry and source-code context to explain the likely cause.

The investigation can include application errors, stack traces, logs, infrastructure telemetry, and relevant repository files. The available evidence determines the context OpsAI can use.

For code-related errors, OpsAI can generate a proposed code change and open a pull request. For supported Kubernetes issues, it can recommend or apply infrastructure-level fixes.

You can investigate issues manually or enable automatic investigation so OpsAI prepares solutions when supported issues are detected.

How Fix with OpsAI Works#

When you start an investigation, OpsAI uses the available context to analyze the issue.

The investigation process includes:

Collect Context — Retrieve relevant metrics, traces, logs, stack traces, and resource information.
Analyze the Issue — Correlate the available telemetry and identify possible causes.
Identify Root Cause — Explain the likely source of the failure.
Generate Solution — Recommend a fix based on the investigation.
Prepare Remediation — Generate a code change or supported infrastructure fix.
Review and Apply — Review the proposed solution and take the appropriate action.

Go to OpsAI Home.

The issue list displays problems detected across your monitored environment.

To investigate an issue:

Locate the issue using the available filters.
Click the issue to open its details.
Review the available error information.
Start an investigation with OpsAI.

Issue Details#

The Issue Details screen provides information about the selected problem.

Depending on the issue type, you can review:

Error message and description
Affected service or resource
Exception and stack trace
Related logs
Available metrics and traces
Issue occurrences
Relevant resource or application context

For application errors, the stack trace can help identify where the failure occurred.

For Kubernetes issues, the related resource information can help identify which workload or configuration is affected.

Investigate an Issue with OpsAI#

From the Issue Details screen, select the option to investigate or fix the issue with OpsAI.

OpsAI analyzes the available telemetry and generates a root cause analysis.

The investigation can include:

What caused the issue
Which service, resource, or code path is affected
Relevant evidence from logs, traces, or error metadata
A recommended solution
A proposed code or infrastructure change, when supported

For example, if an application throws an exception, OpsAI can analyze its stack trace and related telemetry to identify the likely source of the error.

Root Cause Analysis#

OpsAI provides an explanation of the likely cause of the issue based on the available evidence.

For application errors, it can correlate:

Exception stack traces
Error messages
APM traces
Related logs
Historical context
Relevant source code from the connected repository

For infrastructure and Kubernetes issues, it can use resource information, metrics, logs, and other available telemetry.

Use the root cause analysis to understand why the issue occurred before reviewing the proposed fix.

Generate Code Fixes#

When an issue is related to application code, OpsAI can prepare a code-level solution.

To use this capability, connect your source code repository through a supported GitHub or Bitbucket integration.

Once connected, OpsAI can retrieve the files relevant to the error and identify the likely code location responsible for the failure.

The generated solution can include:

Explanation of the required change
Relevant source files
Proposed code modifications
Side-by-side code diff
Option to create a pull request

Create a Pull Request#

After OpsAI generates a proposed code fix:

Review the root cause analysis.
Inspect the suggested code changes.
Review the side-by-side diff.
Use the available option to create a pull request.
Review and merge the pull request through your repository's normal development workflow.

A proposed pull request gives developers a reviewable change linked to the investigation, including the relevant files and an explanation of the fix.

Note: A generated pull request is a proposed fix, not a guarantee that the issue is resolved. Review and test the changes before merging or deploying them.

Fix Kubernetes Issues#

OpsAI can also investigate Kubernetes issues and recommend infrastructure-level fixes.

For supported issues, OpsAI can apply remediation directly to the affected Kubernetes resource without creating a code pull request.

For example, when a workload is failing because of a Kubernetes configuration problem, OpsAI can analyze the available resource data and propose an appropriate configuration change.

Apply a Kubernetes Fix#

Open the Kubernetes issue from OpsAI Home.
Review the issue details.
Start the OpsAI investigation.
Review the root cause analysis and proposed remediation.
If direct remediation is supported, review the change and select the available fix action.
Verify the affected resource after the change is applied.

Kubernetes remediation requires the necessary write permissions. OpsAI can apply changes only when access to the required tools has been explicitly granted.

Fix Vercel Application Issues#

OpsAI can detect and investigate application issues from Vercel using the Middleware Vercel Integration.

Once the integration is configured, OpsAI automatically collects Error and Fatal issues from your Vercel logs and groups related errors to simplify investigation.

Investigate Vercel Issues#

Configure the Vercel Integration.
Open OpsAI Home to view detected Vercel issues.
Use the issue list and filters to find the relevant error.
Click an issue to open its details.
Run Root Cause Analysis (RCA) to investigate the problem.
Review the root cause and suggested solution.

OpsAI uses correlated telemetry and available code context to identify the likely cause of the error.

For code-related issues, OpsAI can also generate a proposed fix and create a pull request for review.

For complete setup and investigation instructions, see Fix Issues with OpsAI — Vercel Integration.

Fix Performance Issues#

OpsAI automatically detects application performance issues from APM and RUM traces, helping you identify slow operations and inefficient resource usage.

It can identify performance problems such as:

N+1 Database Queries — Repeated database queries that increase execution time.
Consecutive Database Queries — Sequential database operations that contribute to latency.
N+1 API Calls — Repeated API requests that increase application response time.
Consecutive HTTP Calls — Sequential HTTP requests that slow down application execution.
Large Frontend Assets — Heavy frontend resources that affect page performance.

Investigate Performance Issues#

Go to OpsAI Home.
Filter the issue list by Performance type.
Select a performance issue to open its details.
Review the affected operation and the time spent executing it.
Start an OpsAI investigation to analyze the issue.
Review the root cause analysis and suggested fix.

OpsAI uses the available APM or RUM telemetry to identify the cause of the performance issue and recommend improvements.

For details about each issue type, see Performance Issues.

Investigate Alerts with OpsAI#

OpsAI lets you centralize and investigate alerts from Middleware, Grafana, and Datadog in one place.

Once the relevant integrations are configured, alerts appear in the OpsAI alerts listing, where you can investigate them using the available observability data.

Investigate an Alert#

Configure the relevant alert integration in Middleware.
Open OpsAI Home and navigate to the alerts listing.
Select an alert to view its details.
Start an OpsAI investigation.
Review the root cause analysis and suggested solution.

OpsAI analyzes the available telemetry associated with the alert to identify possible causes and recommend a fix when possible.

This helps you investigate alerts without manually correlating logs, metrics, and traces across multiple monitoring tools.

For integration and configuration instructions, see OpsAI Settings.

Ask Follow-Up Questions#

After OpsAI generates its analysis or proposed fix, you can continue the investigation through chat.

For example:

Why did this error occur?
Which code change caused the issue?
Is this issue affecting other services?
Can you explain the proposed fix?
What happens if I apply this change?
Is there another way to resolve this issue?

OpsAI uses the existing investigation context to answer follow-up questions.

This is useful when you want to understand the root cause or review a proposed solution before taking action.